近兩年來爆發的新冠肺炎（COVID-19）使人們人心惶惶，無論從事任何活動都應該遵守社交距離以及配戴口罩，所以我們設計了一個可以監測人們是否有保持良好的社交距離以及配戴口罩的程式，從而降低傳播的機率。在監測配戴口罩的部份，我們先利用OpenPose偵測出人體之骨架，然後再利用骨架中之頭部關節擷取出受監測人的臉部區域，最後藉由卷積式類神經網路（Convolutional Neural Networks，CNN）模型架構來實現臉部的影像辨識，透過大量的口罩影像資料和攝影機與電腦程式連線，以此為基礎來訓練分辨人們是否有配戴口罩。在社交距離的部分，我們建立了一個場景影像與實際監控環境間之射影變換，藉此估算影像中人與人之間的距離，有了這個矩陣，我們便能判斷人們是否有保持良好的社交距離。論文的最後我們並設計了一個GUI介面以達成使用上的方便性。

In the last two years, the outbreak of COVID-19 has caused people to get into panic. No matter what activities they are engaged in, they should obey social distancing and wear masks. Therefore, we designed a system that can monitor whether people wear masks and maintain good social distance. The program thereby reducing the probability of transmission. In the part of monitoring wearing masks, we first use OpenPose to detect the skeleton of the human body. We use the head joints in the skeleton to extract the face area of the monitored person. Finally, we use the Convolutional Neural Networks (CNNs) to achieve facial image recognition. By using a large amount of mask image data and camera connection with computer programs, we train a CNN to distinguish whether people wear masks. In the part of social distance estimation, a projective transformation matrix between the scene image and the actual monitoring environment is established to estimate the distance between people in the image. With this matrix, we can judge whether people maintain good social distance. At the end of the research, we build up a GUI to make it convenient for the people.